The National Digital Stewardship Alliance Content Working Group is excited to launch a new series for the Signal. We’re calling it Content Matters, and much like the Insights series that the Innovation Working Group has published here for few years now, the interview series will feature discussions and stories about the content that our members are passionate about, the importance of what they are stewarding and preserving, and the users of this content.
In this inaugural installment we’re excited to chat with Deb Boyer, the project manager for PhillyHistory.org, which provides public access to digitized historic photographs and maps of Philadelphia. Deb was a big part of our Why Digital Maps Can Reboot Cultural History panel at the South By Southwest 2013 conference and brings a unique background to her work with digital maps and cultural heritage.
Butch: Tell us briefly about your background and how you ended up doing what you’re doing.
Deb: I completed my Master’s in Public History at Loyola University Chicago before ending up in Philadelphia a few years ago. At that time the company I work for, Azavea, was looking for someone to manage the PhillyHistory.org website and digitization initiative they developed for the Philadelphia Department of Records.
Our mission focuses on applying geographic data analysis and visualization to improving communities and advancing the state-of-the-art through research. With this focus in mind, we work on a lot of projects related to digital humanities, urban forestry, land conservation, urban planning, elections and other areas where digital mapping and spatial analysis can assist in getting the public involved in creating healthier and more civically engaged communities.
Public historians are all about involving diverse audiences in the exploration of history so I was intrigued at the idea of a digital project developed to provide greater access to historical information.
Deb: PhillyHistory.org provides free access to photographs and maps from the Philadelphia City Archives and other cultural organizations. Users can search over 100,000 images by geographic location (address, intersection, or neighborhood), keyword, and other criteria. We’re still actively scanning images and add new photos to the site every week.
Part of PhillyHistory.org’s uniqueness is simply the age of the project. The site was launched in 2005, and we’ve been adding photos and maps to the website for eight years now. From the beginning, when online mapping tools were just making their appearance, we thought it was important to make the images and maps not only visible online but also searchable by geographic criteria in a way that the general public could understand.
When presented with a large collection of historical materials, people often seek that connection to their personal history – photos of their grandpa’s home or their neighborhood or where their mom went to school. By adding an address and neighborhood search, we wanted to tap into that connection people feel to their local and personal history. Digital and physical preservation of the materials is important, of course, but for us, it always comes down to helping people understand their history from a spatial point of view.
A mobile version of the web site was developed within a few months of the first iPhone release, and in 2010, we tested the feasibility of using the sensors in smart phones to overlay historic images on a view of the contemporary city as seen through the camera of the phone. We wrote up the experiment with augmented reality in a white paper. I think this attention to both practical applications and research is unusual for a municipal archives.
Butch: Mapmaking is really about storytelling across geographies. What attracted you to use maps as your storytelling mode (as opposed to video, audio, other interactive, etc.)?
Deb: The main source of materials for PhillyHistory.org is the photographic collection of the Philadelphia City Archives. Estimated at over two million images, the collection is unique in that the images are of specific places. Rather than a lot of images of people or events, we have photo after photo of houses and streets – many with meticulously documented locations and covering over 150 years of Philadelphia history.
The location of the photos was an important reason why they were made in the first place. Many of the photos originally were taken as a risk management tool to document public works projects as Philadelphia grew and changed. Geographic context is such an important element of the photos, and we felt that tying the images to their location makes it easier for users to understand how the photos tell the story of a location.
Neighborhoods are also extremely important to Philadelphians. People feel strongly about their local communities, and the location-based search is by far the most frequently used search criteria on PhillyHistory. Often, people connect their personal stories and histories to specific places and the businesses, places of worship, schools, and events that existed or occurred at those locations. By placing the images on a map interface and by displaying historic maps, we hope to encourage users to explore both the history of a neighborhood and how that history has affected the ongoing change and development of Philadelphia.
Butch: How have the technologies of digital mapping changed over the past five years? How have those changes affected the work you do?
Deb: One of the most significant changes has been the increase in free and open source tools for mapping. It’s so much easier for people without Geographic Information Systems experience to explore spatial analysis and mapping now compared to five years ago. WordPress, Omeka, and other content management sites used by many Libraries, Archives and Museums (LAMs) now have great add-ons for including geography on websites or in exhibits. Having a geospatial component has become the norm.
Five years ago, I had to spend ten minutes explaining what it was we were trying to do with geography on PhillyHistory. Now, I just have to say that you can search the photos by location and nearly everyone understands. I think people are experimenting with using place and space in new ways for both digital storytelling and understanding the past. It’s a wonderful change.
In just the last couple years, we’ve also seen an increase in web-based software for geographic analysis. People are beginning to think beyond just displaying points on a map and are increasingly thinking about how the data can be transformed with analytical and statistical tools. Until fairly recently, this kind of analytical capability has largely been limited to desktop software.
With new cloud computing platforms it is becoming possible to do this kind of work on the web. Our recent research has been focused on developing an open source toolkit, called GeoTrellis, that can break geoprocessing tasks into many concurrent units of work, generate results, and return interactive geospatial data in less than a second. Faster results mean that we can create new kinds of user experiences.
We used GeoTrellis when we worked with the Digital Scholarship Lab at the University of Richmond to create the Visualizing Emancipation website, which enables users to explore various events in the emancipation process during the Civil War. With a click of a button, users can create heat maps, view spatial data animations, and search and display layers of data related to the legality of slavery, Union Army troop locations, railroads, and emancipation events. It’s all generated on the fly, and it would have been extremely difficult to do five years ago.
While digital mapping is becoming much more common in technology projects, we also need to remember that people may still have difficulty reading maps and understanding things spatially. Many individuals may also primarily access the internet through mobile and tablet devices. It’s easy to get excited about the great advances in digital mapping, but it’s also crucial to remember that the maps must be presented in an intuitive way that provides a good user experience across multiple devices.
Butch: Where are the current gaps in terms of tools and services to help digital storytellers do their work with maps? What are some tools, approaches or initiatives that might remake the future landscape of digital mapping?
Deb: We’ve long been interested in and struggled with the concept of historic place names. Over the years of a city’s life, urban planners and city officials change street names, reroute roads, and renumber house addresses. The general public, meanwhile, may refer to roads, buildings, and neighborhoods by the local, unofficial name that never shows up on any map or may adopt a new name for a place and quickly forget the old.
How do we map the past? How do we geocode a photo when the road where that photo was taken no longer exists or the contemporary name is different from the historic one? Expecting LAM employees to know and identify the current latitude and longitude coordinates for every street, address, building, temple, theater, or other location that ever existed is simply impossible. While there have been digital gazetteers that translate place names, much of our historical material is much more specific. We really need a geographic database that combines elements of a historic gazetteer with a digital version of a historic street map.
For the PhillyHistory project, we created the Philadelphia Historic Streets Index to track changes in street names from 1858 to the present. It’s a good first step for Philly, but we need a system that’s both more refined at identifying particular street segments and broad enough to apply to other cities. We recently won a National Science Foundation Small Business Innovation Research (SBIR) Grant to explore the creation of a historic street and address database plus a temporal geocoder.
The prototype will leverage the crowdsourcing approach that has been so successful for projects like OpenStreetMap to connect historic geographic materials with contemporary places. We are collaborating with a group at New York Public Library and the OpenStreetMap project also has a related effort. Approaching this issue from several different angles will hopefully result in a breakthrough that makes the mapping of historic locations much more efficient for cultural institutions that might not have as much experience with geographic software.
Butch: At NDIIPP we’ve started to think more about “access” as a driver for the preservation of digital materials. To what extent do preservation considerations come into play with the work that you do? How does the provision of enhanced access support the long-term preservation of digital geospatial information?
Deb: For PhillyHistory, providing access to the maps and photos meant that we had to create a digital preservation plan. Access and preservation were intertwined from the very beginning. Prior to the PhillyHistory project, access to the photographic collection had been limited in order to preserve the collection, but even with limited access, the collection was deteriorating and poorly organized.
Digitization was a path to three related objectives: improve the ability to find relevant material, improve preservation by moving negatives to more stable storage environments following scanning, and dramatically increase access by enabling the general public to use the materials in an online application. Improved access has also helped to fund the preservation effort. Licensing of high resolution photos, sales of prints, and other revenue have helped to both increase usage of the material and generate revenue for the Archives. The additional revenue does not cover all of the digitization cost, but it does help to support the effort.
What I find most interesting in the interaction between enhanced access and preservation is the idea of how to preserve the geospatial data created on many of the new interactive mapping projects. When an organization chooses to display digital materials on a digital map or enables their users to create their own maps connecting digital items, how are those new maps preserved?
In the Visualizing Emancipation project, the map is a huge part of telling the story of the emancipation data contained on the website. Users interact with the map and create entirely new maps by their search criteria, animation choices, and selected filters. Those new maps are inherently ephemeral and change the moment that another search selection is made or new data is uploaded. How do we capture the transient items created in web-based digital mapping?
Software companies often focus on backing up the code and database and LAM organizations think of backing up the original data sources (both digital and physical). Are screenshots or the ability to export a map as a PDF sufficient measures to capture the new maps being created as part of our interactive projects? Do we need to preserve these items? At first, it seems like all the hundreds of variations of digital maps that could be created are superfluous things to save.
Many LAMs, however, are trying to preserve the comments, stories, and other digital interactions generated by public interaction with the LAM collections via social media, online exhibitions, and other digital projects. Why should digital geospatial data be any different?
Butch: In the course of your workflow you incorporate already extant digital materials (images, documents, etc.) as well as create your own (digital maps). What role does your project take in the long-term stewardship of any of these materials? How does the work you do help secure the long-term stewardship of digital geospatial data?
Deb: For the PhillyHistory.org project, our digitization process includes creating three digital versions of each image that are then stored in three separate locations. The server that houses one of those versions is then backed up again and stored in a separate location. As is traditional, the physical negatives are rehoused in archival quality enclosures and stored at the City Archives, but we think a long-term stewardship plan for the digital files is just as important. We, of course, regularly backup the code and databases that support the digital projects.
Digital humanities software projects, in many ways, help secure the long-term stewardship of digital geospatial data by spreading awareness of the existence of that data. When the historic maps are visible and searchable online, rather than only accessible in the depths of an archives, a broader audience that goes beyond the map enthusiasts becomes aware of the resources and the necessity of their preservation. A map might not be available for decades but put it online, build an audience who uses it, and then watch how quickly people become annoyed when that mapping website or project is no longer available. When organizations start to think about making resources available online, they have to think about the long-term stewardship of that digital data.
Butch: How widespread is an awareness of digital stewardship and preservation issues in the part of the geographic community in which you operate?
Deb: I think that many companies that create software are aware of the need to backup and preserve the code they write. Most projects will have a system of built-in backups of the database in case of unexpected failures. Long-term archiving of that code and providing access to it is much more complex. Most open source projects use a fairly sophisticated versioning system, such as GitHub or Subversion, that saves all previous versions of the source code and enables developers to view the code at various points in its history as well as track when changes are made.
For our spatial analysis work, the word “metadata” is nearly as important as it is in the LAM community as an element of long-term stewardship of data. Technical metadata ensures that the geospatial information can be reused in another format and understood by other spatial analysts, and the geospatial community has generally paid far more attention to metadata standards than most other technology communities.